Adaptyv (Claude Skill)

Claude skill that closes the design–build–test loop for proteins and antibodies — pre-screens sequences in silico (solubility, expression, stability, interface, aggregation), then submits them to the Adaptyv cloud laboratory for wet-lab binding, expression, thermostability, and enzyme-activity assays.

   
Type Claude Skill
Supplier K-Dense Inc.
Availability GA — distributed via the K-Dense scientific-agent-skills collection. The Adaptyv platform itself is in alpha/beta and API access is request-gated by Adaptyv.
Pricing Free / OSS (skill source). Adaptyv wet-lab assays are paid per-experiment via Adaptyv; pricing is quoted by Adaptyv on API onboarding.
Capabilities Read/Write — local in-silico pre-screening, plus authenticated API calls that submit experiments to a wet lab (the skill writes to the Adaptyv platform on the user’s behalf).

How to install

  • Also packaged in the SciAgent-Skills collection (jaechang-hits (community OSS, CC BY 4.0)): clone jaechang-hits/SciAgent-Skills and run /plugin install sciagent-skills in Claude Code (or copy skills/proteomics-protein-engineering/adaptyv-bio into ~/.claude/skills/).
  • Claude Code / Claude.ai — Skills CLI (recommended): 
    npx skills add K-Dense-AI/scientific-agent-skills

    Installs the K-Dense collection; enable the adaptyv skill when prompted (also works in Cursor/Codex via the Agent Skills spec; requires Node ≥ 18). This installs the full K-Dense scientific-skills plugin, which bundles the adaptyv skill alongside ~170 sibling skills.

  • Claude Code / Claude Desktop — manual clone (single-skill install): 
    git clone https://github.com/K-Dense-AI/scientific-agent-skills
cp -r scientific-agent-skills/skills/adaptyv ~/.claude/skills/

    (Replace ~/.claude/skills/ with your user-level Claude skills directory if you’ve relocated it.) The skill uses uv to install its Python dependencies on first invocation; install uv first (pipx install uv or per the uv install docs) so the skill can resolve its environment.

  • Authentication — Adaptyv requires an API key. Email [email protected] to request access while the platform is in alpha/beta, then set: 
    export ADAPTYV_API_KEY=<your-key>

    in the environment Claude Code (or Claude Desktop) launches the skill from.

What it does

  • Submits and tracks wet-lab experiments on the Adaptyv cloud platform via authenticated API calls:
    • Binding assays — biolayer interferometry for protein–target K_D / k_on / k_off
    • Expression testing — E. coli, mammalian, yeast, or insect cell systems
    • Thermostability — DSF and CD for T_m and stability profiling
    • Enzyme activity — kinetic parameters, substrate specificity, inhibitor screens
  • Pre-screens sequences in silico before submission so users don’t waste wet-lab cycles on candidates that will fail expression or solubility:
    • NetSolP / SoluProt — solubility prediction
    • SolubleMPNN — sequence redesign to improve expression
    • ESM — sequence-likelihood scoring
    • ipTM (AlphaFold-Multimer) — interface-stability estimate for designed binders
    • pSAE — aggregation-risk quantification
  • Supports batch submission for library screening and webhook callbacks for experiment completion (Adaptyv turnaround is ~21 days).
  • Returns kinetic parameters, confidence metrics, and raw assay data alongside the original sequence metadata so subsequent design iterations can be conditioned on real wet-lab feedback.

Primary use cases: antibody affinity maturation, therapeutic-protein developability assessment, enzyme engineering, AI-driven protein-design validation loops, library-scale expression screens, lead optimization with experimental ground truth.

Notes

  • The skill writes to a wet lab. Review submitted sequences and experiment configurations carefully before approval — each call spends Adaptyv credits and consumes platform-side wet-lab resources.
  • Adaptyv API is gated. Access is alpha/beta as of 2026-06-02; without an issued API key the in-silico pre-screening half is still usable, but the submission tools will fail.
  • K-Dense skill path migration. The canonical K-Dense scientific-agent-skills repo moved skill directories from skills/<name>/ to skills/<name>/ in v2.43.0 to match the Agent Skills CLI layout. The published marketplace repo (K-Dense-AI/scientific-agent-skills) still uses skills/<name>/ paths at the time of verification; the manual-clone snippet above matches that layout. If the marketplace path migrates, update the cp source path accordingly.
  • Pairs with the AlphaFold (Claude Skill), PDB (Claude Skill), and Glycoengineering (Claude Skill) entries for an antibody-design workflow that runs from sequence → structure → glycan profile → in-silico screen → wet-lab validation.

Sources

Installed this tool?

Share feedback — install path, OS, errors, workarounds. The form opens with this tool pre-selected and a link back to this page.